-
Notifications
You must be signed in to change notification settings - Fork 9.4k
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Signed-off-by: Jia-wei Yu <yoga.yu.169@gmail.com>
- Loading branch information
Showing
484 changed files
with
21,623 additions
and
11,429 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,44 @@ | ||
# Copyright VMware, Inc. | ||
# SPDX-License-Identifier: APACHE-2.0 | ||
|
||
{{- $nodes := .Vars.worker.replicaCount }} | ||
file: | ||
/job/hostfile: | ||
exists: true | ||
contains: | ||
{{- range $i, $e := until $nodes }} | ||
- /deepspeed-worker-{{ $i }}/ | ||
{{ end }} | ||
{{ .Vars.client.persistence.mountPath }}: | ||
exists: true | ||
filetype: directory | ||
mode: "2775" | ||
owner: root | ||
addr: | ||
{{ range $e, $i := until $nodes }} | ||
tcp://deepspeed-worker-{{ $i }}-external:{{ $.Vars.worker.externalAccess.service.ports.ssh }}: | ||
reachable: true | ||
timeout: 500 | ||
tcp://deepspeed-worker-{{ $i }}.deepspeed-worker-headless:{{ $.Vars.worker.containerPorts.ssh }}: | ||
reachable: true | ||
timeout: 500 | ||
{{ end }} | ||
command: | ||
run-git-example: | ||
exec: bash -c ". /opt/bitnami/scripts/deepspeed/entrypoint.sh && cd /app/benchmarks/communication/ && deepspeed all_reduce.py --mem-factor 0.0001" | ||
exit-status: 0 | ||
timeout: 900000 | ||
{{- $uid := .Vars.client.containerSecurityContext.runAsUser }} | ||
{{- $gid := .Vars.client.podSecurityContext.fsGroup }} | ||
check-user-info: | ||
# The UID and GID should always be either the one specified as vars (always a bigger number that the default) | ||
# or the one randomly defined by openshift (larger values). Othenrwise, the chart is still using the default value. | ||
exec: if [ $(id -u) -lt {{ $uid }} ] || [ $(id -G | awk '{print $2}') -lt {{ $gid }} ]; then exit 1; fi | ||
exit-status: 0 | ||
{{ if .Vars.client.serviceAccount.automountServiceAccountToken }} | ||
check-sa: | ||
exec: cat /var/run/secrets/kubernetes.io/serviceaccount/token | cut -d '.' -f 2 | xargs -I '{}' echo '{}====' | fold -w 4 | sed '$ d' | tr -d '\n' | base64 -d | ||
exit-status: 0 | ||
stdout: | ||
- /serviceaccount.*name.*deepspeed-client/ | ||
{{ end }} |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,37 @@ | ||
source: | ||
type: "git" | ||
git: | ||
repository: https://github.com/microsoft/DeepSpeedExamples | ||
revision: master | ||
client: | ||
podSecurityContext: | ||
enabled: true | ||
fsGroup: 1002 | ||
containerSecurityContext: | ||
enabled: true | ||
runAsUser: 1002 | ||
runAsGroup: 1002 | ||
serviceAccount: | ||
create: true | ||
automountServiceAccountToken: true | ||
persistence: | ||
enabled: true | ||
mountPath: /bitnami/deepspeed/vib-test | ||
worker: | ||
replicaCount: 2 | ||
containerPorts: | ||
ssh: 3333 | ||
# We found an issue when the worker belonged to a group | ||
# different from 1001 | ||
podSecurityContext: | ||
enabled: true | ||
fsGroup: 1004 | ||
containerSecurityContext: | ||
enabled: true | ||
runAsUser: 1004 | ||
runAsGroup: 1004 | ||
externalAccess: | ||
enabled: true | ||
service: | ||
ports: | ||
ssh: 4444 |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,68 @@ | ||
{ | ||
"phases": { | ||
"package": { | ||
"context": { | ||
"resources": { | ||
"url": "{SHA_ARCHIVE}", | ||
"path": "/bitnami/deepspeed" | ||
} | ||
}, | ||
"actions": [ | ||
{ | ||
"action_id": "helm-package" | ||
}, | ||
{ | ||
"action_id": "helm-lint" | ||
} | ||
] | ||
}, | ||
"verify": { | ||
"context": { | ||
"resources": { | ||
"url": "{SHA_ARCHIVE}", | ||
"path": "/bitnami/deepspeed" | ||
}, | ||
"target_platform": { | ||
"target_platform_id": "{VIB_ENV_TARGET_PLATFORM}", | ||
"size": { | ||
"name": "L4" | ||
} | ||
} | ||
}, | ||
"actions": [ | ||
{ | ||
"action_id": "goss", | ||
"params": { | ||
"resources": { | ||
"path": "/.vib" | ||
}, | ||
"tests_file": "deepspeed/goss/goss.yaml", | ||
"vars_file": "deepspeed/runtime-parameters.yaml", | ||
"remote": { | ||
"pod": { | ||
"workload": "deploy-deepspeed-client" | ||
} | ||
} | ||
} | ||
} | ||
] | ||
}, | ||
"publish": { | ||
"actions": [ | ||
{ | ||
"action_id": "helm-publish", | ||
"params": { | ||
"repository": { | ||
"kind": "S3", | ||
"url": "{VIB_ENV_S3_URL}", | ||
"authn": { | ||
"access_key_id": "{VIB_ENV_S3_ACCESS_KEY_ID}", | ||
"secret_access_key": "{VIB_ENV_S3_SECRET_ACCESS_KEY}" | ||
} | ||
} | ||
} | ||
} | ||
] | ||
} | ||
} | ||
} |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,51 @@ | ||
{ | ||
"phases": { | ||
"package": { | ||
"context": { | ||
"resources": { | ||
"url": "{SHA_ARCHIVE}", | ||
"path": "/bitnami/deepspeed" | ||
} | ||
}, | ||
"actions": [ | ||
{ | ||
"action_id": "helm-package" | ||
}, | ||
{ | ||
"action_id": "helm-lint" | ||
} | ||
] | ||
}, | ||
"verify": { | ||
"context": { | ||
"resources": { | ||
"url": "{SHA_ARCHIVE}", | ||
"path": "/bitnami/deepspeed" | ||
}, | ||
"target_platform": { | ||
"target_platform_id": "{VIB_ENV_TARGET_PLATFORM}", | ||
"size": { | ||
"name": "L4" | ||
} | ||
} | ||
}, | ||
"actions": [ | ||
{ | ||
"action_id": "goss", | ||
"params": { | ||
"resources": { | ||
"path": "/.vib" | ||
}, | ||
"tests_file": "deepspeed/goss/goss.yaml", | ||
"vars_file": "deepspeed/runtime-parameters.yaml", | ||
"remote": { | ||
"pod": { | ||
"workload": "deploy-deepspeed-client" | ||
} | ||
} | ||
} | ||
} | ||
] | ||
} | ||
} | ||
} |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Oops, something went wrong.